This paper presents amethod to design a cost-optimal nonredundant sensor network to observe all variables in a general nonlinear process. Arnixed integer linear programming model was used to minimize the cost with data classification to check the observability of variables. This work is a starting point for designing sensor networks for general nonlinear processes based on various criteria, such as reliability and accuracy.
Published in:
Tsinghua Science and Technology
(Volume:8
,
Issue:
6
)
Date of Publication: Dec. 2003